A collection of academics, engineers, and industry professionals gathered recently on the MIT campus to discuss the present state of the Internet of Things and consider how best to create a promising future for the technology. This blog highlights some of the key ideas that were shared during an IoT Thought Leaders Panel.
On October 6-8th, MIT Media Labs hosted the Fourth International Conference on the Internet of Things. The event gathered academic and industry experts from around the world, and included a number of technical workshops, inspiring keynotes, and demonstrations of some of the leading-edge research and developments around IoT applications, systems, and architectures. During the event, I sat in on a panel of six experts in the IoT field who discussed their views on what it will take to bring IoT to the next level. The panel included three academic professionals and three industry experts and brought to light some of the most pressing challenges with IoT as it is currently being adopted as well as some of the greatest opportunities this technology offers.
A Challenging Landscape for Success of IoT
The panel opened by considering some of the main challenges facing IoT development and adoption. Among the key issues that were discussed are the following, in order of resonance across the panel:
- Security - this was by far the most repeated challenge that panelists see standing in the way of widespread, successful IoT adoption. One of the panelists shared the analogy that security used to be physical, with systems and devices that were physically protected from a network or outside intrusion by nature of their physical distance (air gaps). That has changed immensely with the Internet and online connectivity. Panelists offered additional examples of security challenges around upgrades to connected devices. With billions of IoT enabled devices across vast, physically separated networks today and hundreds of billions expected in coming years, it is quite a challenge to make secure software updates or modernize older devices.
- Lack of a Dominant Architecture - one of the speakers, MIT Professor Sanjay Sarma, cautioned against the ad hoc implementations of today around IoT. He shared the history of successful industries being formed around dominant architectures, including the important but counterintuitive decision to base electricity on AC rather than DC. Prof. Sarma advocated that a dominant architecture is critical to moving into the next phase of IoT. Building on the architecture theme, Japan's Keio University Professor Jun Murai added the importance of having a shared architecture that would span industries and information systems. Attendees and panelists alike agreed that, similar to RFID, IoT is in need of a common, open, standards-based architecture to successfully pass into the next phase of widespread adoption.
- Future-proofing IoT - this challenge relates very closely with the manufacturing industry because it addresses the types of devices or assets that are and will be connected through IoT. While much of the thinking today is related to smartphones and tablet, the panelists cautioned the shortsightedness of this approach to developing standards and an architecture to support them. Rather, the panel encouraged the wider audience to consider all of the much longer-lived devices (or assets) that will be connected. Take, for example, an airplane or its engine - these are assets that survive decades, and their ongoing connectivity via IoT will be imperative for value realization from IoT. Therefore, panelists urged the engineers and developers in the audience to consider developing business models and architectures that do not break down with scale.
Despite these identified challenges, all of the panelists expressed positive views about the transformational potential that IoT brings to industry. Manufacturing was viewed as an industry that not only stands to gain significant value from IoT, but also offers some best practices to follow around resiliency.
Resiliency for IoT Can Borrow from Manufacturing Playbook
At IDC Manufacturing Insights, we've spent quite some time over the past several years talking about resiliency, which we define in the same way as the academic discipline of resilience: the ability to adapt to changing circumstances while maintaining a central purpose; a key construct is to be simple at the core, diverse at the edges. We have considered resiliency in terms of global manufacturing supply chains and modernizing the factory floor. The industrial automation that has occurred on the factory floor, and the creation of multiple layers of network design in supply chains to support resiliency are both examples of how leading manufacturers understand the need for resilience.
Regarding IoT, thought leaders at the conference discussed the need to add resilience to IoT networks - in the form of creating multi-layered networks that remove single failure points, and accommodate modernizing existing networks to co-exist with the new "webs" of connected "things." Another form of resilience concerns the devices themselves. By creating a network that is de-centralized, multi-layered to avoid single points of failure, and secure, much of the "control" will be pushed out to the edges of the network, to the devices themselves. This opens up the opportunity for connected devices to offer multi-purpose opportunities.
Consider the sensor that is sitting in a building, monitoring the temperature of a room. The majority of this sensor's time is spent in waiting mode. What if that sensor could offer its services to the network for another purpose? What if it could communicate with the other energy-sensor networks in the building to determine spikes in energy consumption across HVAC and electricity? What if this sensor could advertise its availability for use when it is not tracking the temperature? All of these scenarios point toward making the devices resilient on top of a resilient IoT network. And, manufacturers can surely imagine dozens of applications for resilient IoT related to their businesses and supply chains that will provide not only improved efficiencies, but also new business models and ways of monetizing IoT that have yet to be tried.
It is an exciting time for IoT, and I remain optimistic that the industry leaders and academic experts will work together to forge the best architecture to support the full potential that IoT brings. While I am not naïve to believe this will happen overnight, events such as MIT's International Conference on the Internet of Things provide enriching opportunities for leaders to gather and determine the best ways to work toward a common goal of widespread IoT adoption. Leading manufacturers are part of this conversation, and the results will influence the industry's ability to more readily exploit the technology that is top of mind for many executives across manufacturing segments.